Study On Downscaling With Multisource Precipitation Data In Western China

Precipitation is one of the key processes involved in the global water cycle,and it is also an important part of the Earth's surface material energy cycle and exchange process.Therefore,high-precision precipitation data is extremely useful for studying the spatial and temporal distribution of precipitation in a region.The measured data of traditional rainfall stations are discrete point data.While the emerging remote sensing precipitation data are continuous,they are limited by the spatial resolution,and the precipitation distribution in a small local area cannot be accurately represented.Therefore,how to obtain continuous high-precision and high spatial resolution precipitation data through some technical means has become an increasingly important issue in meteorology and geography.On the basis of summarizing predecessors' studies,this paper uses the western region of China as the scope of the study.We first evaluated the three sets of remote sensing precipitation data of CMORPH 1.0(gauge-satellite),TRMM 3b43V7,and CMPA(CMPA-Daily)using conventional indicators and topographic performance.Applicability,and determine the background field of downscaling research;then construct a downscaling model based on longitude,latitude,elevation,slope correction factor,and add vegetation cover index and water vapor content in the two sets of models,and according to the study area The partitions were modeled;in the end,GNSS technology was used to invert the water vapor,and an exploratory attempt was made to further optimize the downscaling model.The following conclusions were obtained:(1)The three sets of precipitation data can,on the whole,accurately reflect the amount of precipitation and distribution.Compared with CMORPH and TRMM,CMPA precipitation data has higher accuracy and better data stability in annual total precipitation and multi-year monthly average precipitation estimates.In complex terrain,the accuracy of CMPA precipitation data is better than that of CMORPH and TRMM.Therefore,the CMPA precipitation data is selected as the background field for downscaling research.(2)Based on the premise of large geographic and climatic differences in the study area,this paper uses the Kunlun Mountains,the A-erh-chin Mountains,and the Qilian Mountains as the dividing line on the northern slope of the Qinghai-Tibet Plateau to build downscale models for the southwest and northwest regions,respectively.Overall,the accuracy of the downscaling model that incorporates the vegetation cover index is better than the downscale model that incorporates the water vapor content.Overall,the accuracy of downscaling results in April,July,and October is relatively good.The accuracy of downscaling results in January is slightly lower than in other months,but it is still superior to the CMPA raw data.(3)The water vapor content of three IGS stations in Lhasa,Kunming,and Urumqi in 2010 was retrieved by GNSS technology.The results showed that the GNSS water vapor content accuracy was better than that of ground station water vapor and MOD05 vapor data.Further,the water vapor data of MOD05 was revised and added to the downscaling model through GNSS vapor data.The results show that this research can effectively improve the accuracy of the downscaling model.